Agent based modeling-presentation

Adam Getchell

Nonlinear Physics: Modeling Chaos and Complexity

What is an Agent?Historically related to the Von Neumann machine, as

later improved by Stanislaw Ulam into the first agent-based device – the cellular automata

Agents have:

Activity

Autonomy

Heterogeneity

Agent Activity Goal-direction

Reactivity/Perceptivity to its surroundings (model)

Mobility: Able to roam the model space independantly

Bounded Rationality (imperfect information)

Interacts/exchanges information with other agents, which may in turn cause:

Adaptation: Change in behavior based on interactions with the model or other agents

What is a Model?import breve

class myControl( breve.Control ):def __init__( self ):

breve.Control.__init__( self )self.walkerShape = NonemyControl.init( self )

def getWalkerShape( self ):return self.walkerShape

def init( self ):print '''Setting up the simulation.'''self.pointCamera( breve.vector( 0, 0, 0 ), breve.vector( 0, 60, 0 ) )self.walkerShape = breve.createInstances( breve.Sphere, 1 ).initWith( 1 )breve.createInstances( breve.RandomWalker, 200 )

breve.myControl = myControlclass RandomWalker( breve.Mobile ):

def __init__( self ):breve.Mobile.__init__( self )RandomWalker.init( self )

def init( self ):self.setShape( self.controller.getWalkerShape() )self.setColor( breve.randomExpression( breve.vector( 1.000000, 1.000000, 1.000000 ) ) )self.move( breve.randomExpression( breve.vector( 0.100000, 0.100000, 0.100000 ) ) )

def iterate( self ):self.setVelocity( ( breve.randomExpression( breve.vector( 60, 60, 60 ) ) - breve.vector( 30, 30, 30 ) ) )

breve.RandomWalker = RandomWalker

# Create an instance of our controller object to initialize the simulation

myControl()

Tools and Languages in 2008Program Language(s) Description

Swarm Objective-C, Java Agent modeling library, dated, last release version 2.2 February 2005

RepastJ Java Based on Swarm, written specifically in Java

RepastPy Python Friendly GUI, uses Python for scripting, limited

Repast.NET C#, VB.NET Leverages .NET framework, doesn’t work with Visual Studio 2008

Repast Simphony Java Full-featured, uses Eclipse IDE, can be difficult to setup

MetaABM Java Full-featured, uses Eclipse IDE plus own GUI, designed to use standard model file that can work with other tools (Repast, Weka, VisAD, MatLAB), can be difficult to setup

Breve Steve, Python, Push Fast, easy to use 3d simulation environment targeted towards physics and artificial life

RepastPy -- Model Simple GUI which generates Java classes

RepastPy – AgentSimple Python scripting for behaviors

Repast Simphony Uses Java + Groovy to compile an application

Boids A kind of 3D Life simulation producing chaotic

behavior. The rules are:

1. Boids try to fly towards the center of mass of neighboring boids (usually, the perceived CoM with respect to that particular boid)

2. Boids try to keep a small distance away from other objects (including other boids)

3. Boids try to match velocity with near boids (perceived velocity of neighbors)

A Simphony of Boids

breve – basic Controller/Agent structure (Python)import breve

class HelloWorld( breve.Control ):def __init__( self ):

breve.Control.__init__( self )

def iterate( self ):print '''Hello, world!'''breve.Control.iterate( self )

breve.HelloWorld = HelloWorld

# Create an instance of our controller object to initialize the simulation

HelloWorld()

breve – basic Controller/Agent structure (steve)@include "Control.tz"

Controller HelloWorld.

Control : HelloWorld {

+ to iterate:

print "Hello, world!".

super iterate.

}

breve – Gravity & 3D collisions

breve – Gray Scott model of reaction diffusion

breve – Capture the Flag

breve – boids to evolving swarms

Evolving swarmsIn addition to the behaviors of boids, swarm agents:

Seek out food, which randomly teleports around

Feed their friends with excess food

Reproduce when energy (food) hits certain threshold

Die when they run out of energy, or reach maximum age

Land on the ground, rest, fly around again

Mutate in such a way as to improve/reduce reproduction

So how to you mutate code that must be pre-defined?

PushGenetic programming – random crossover and mutation

of computer programs

Doesn’t work for most computer languages, since they typically have rigid syntax:

This makes sense:

L = [math.exp(val) for val in eigenvalues]

This does not:

eigenvalues ] in math.exp(val) = L ] for

Push Programs made up of: instructions, literals, and

sublists

Push program is an expression, entirely placed on the stack and evaluated recursively according to these rules:

1. If P is an instruction then execute it

2. Else if P is a literal then push it on to the stack

3. Else (P must be a list) sequentially execute each of the Push programs in P

Sample Push program and execution( 2 3 INTEGER. * 4.1 5.2 FLOAT.+ TRUE FALSE BOOLEAN.OR )

Pushing onto the stack from left to right, we then pop the stack right to left :

First run is: BOOLEAN.OR FALSE TRUE = (TRUE) (BOOLEAN stack)

Next we have: FLOAT.+ 5.2 4.1 = (9.3) (FLOAT stack)

Finally we have INTEGER.* 2 3 = (6) (INTEGER stack)

Note that each stack has its own type, the stack-based typing system puts each instruction on its own type of stack, so that any combination remains semantically valid. We could re-order all of these stacks without issue.

The main trick is to devise programs that actually produce changeable behaviors in the agents, so they can be selected for or against

References “Agent-based model,” Wikipedia, the free encyclopedia,

http://en.wikipedia.org/wiki/Agent_based_modeling. Christian Castle and Andrew Crooks, Principles and Concepts of Agent-Based Modelling

for Developing Geospatial Simulations, UCL Working Papers Series (UCL Centre for Advanced Spatial Analysis, September 2006), http://www.casa.ucl.ac.uk/working_papers/paper110.pdf.

“Main Page - SwarmWiki,” http://www.swarm.org/index.php?title=Main_Page. Repast Agent Simulation Toolkit, http://repast.sourceforge.net/. Miles Parker, metaABM, http://metaabm.org/docs/index.html. Jon Klein, breve: a 3d Simulation Environment for Multi-Agent Simulations and Artificial

Life (Hampshire College), http://www.spiderland.org/. Craig Reynolds, “Boids (Flocks, Herds, and Schools: a Distributed Behavioral Model),”

Boids, Backround and Update, http://www.red3d.com/cwr/boids/. Abelson et al., “Gray Scott Home Page,” Gray Scott Model of Reaction Diffusion,

http://www.swiss.ai.mit.edu/projects/amorphous/GrayScott/. Jon Klein, “"Push": a Language for Evolutionary Computation Integrated With breve |

breve,” http://www.spiderland.org/node/2759. Lee Spector, Push, PushGP, and Pushpop (School of Cognitive Science: Hampshire

College), http://hampshire.edu/lspector/push.html.